Social Support’s Relationship to Correctional Staff Job Stress,Job Involvement, Job Satisfaction, And Organizationalcommitment

8/18/2019 Social Support’s Relationship to Correctional Staff Job Stress,Job Involvement, Job Satisfaction, And Organizationalc…

1/11

Please cite this article in press as: Lambert, E. G., et al. Social support’s relationship to correctional staff

job stress, job involvement, job satisfaction, and organizational commitment. The Social Science Journal (2015),

http://dx.doi.org/10.1016/j.soscij.2015.10.001

ARTICLE IN PRESSG Model

SOCSCI-1290; No.of Pages11

The Social Science Journal xxx (2015) xxx–xxx

Contents lists available at ScienceDirect

The Social Science Journal

j ourna l homepage: www.e lsev ier .com/locate /sosc i j

Social support’s relationship to correctional staff job stress,

job involvement, job satisfaction, and organizational

commitment

Eric G. Lambert a,∗, Kevin I. Minor b,1, James B. Wells b,2, Nancy L. Hogan c,3

a Department of Legal Studies, TheUniversity of Mississippi, 202 Odom Hall, Mississippi, MS 38677, USAb School of Justice Studies, Eastern Kentucky University, Stratton 467, 521 Lancaster Avenue, Richmond, KY 40475, USAc School of Criminal Justice, Ferris State University, 525 BishopHall, Big Rapids, MI 49307, USA

a r t i c l e i n f o

Article history:

Received 18 December 2014

Received in revised form 18 April 2015

Accepted 5 October 2015

Available online xxx

Keywords:

Correctional staff

Social support

Job stress

Job satisfaction Job involvement

Organizational commitment

a b s t r a c t

The literature suggests that social support, in general, is linked to positive outcomes among

correctional staff, but the different types of social support may differ in their effects. Using

survey data from staff working at a privately-owned,maximum security prison for juveniles

sentenced as adults, this study analyzes three intra-organizational supports, administra-

tive, supervisory, and coworker, andone extra-organizational form, family/friends support,

as antecedents of job stress, job involvement, job satisfaction, and organizational com-

mitment. Personal variables serve as controls. Administrative and supervisory support are

inversely related to job stress, while job involvement is affected positively by supervi-

sory support and negatively by family/friends support. All three intra-organizational forms

of support are significant antecedents of job satisfaction and organizational commitment

however family/friends support is not. These findings hold implications for improving job

outcomes among correctional staff and for future research.

© 2015 Western Social Science Association. Published by Elsevier Inc. All rights reserved.

Prisons are expensive to administer and operate.

Henrichson and Delaney (2012) report that over 35 billion

dollars is spent on prisons each year in the United States

to house 1.4 million adult inmates. Operating prisons is

labor intensive, with staff being the largest expenditure,

often accounting for over 75% of a prison’s annual budget

(Camp & Lambert, 2005). Additionally, staff are a valu-

able resource, as they are responsible for the multitude of tasks and duties necessary for the operation of a humane,

∗ Corresponding author. Tel.: +1 662 915 2672.

E-mail addresses: [email protected] (E.G. Lambert),

[email protected] (K.I. Minor), [email protected] (J.B. Wells),

[email protected] (N.L. Hogan).1 Tel.: +1 859 622 2240.2 Tel.: +1 859 622 1158.3 Tel.: +1 231 591 2664.

secure, and safe prison. In a sense, staff are a prison’s heart

and soul. In an era of reduced government budgets, inves-

tigating the factors that may affect correctional staff is

imperative.

A growing body of research focuses on prison staff,

particularly in terms of job stress, job involvement, job

satisfaction, and organizational commitment. Stressed,

uninvolved, dissatisfied, and uncommitted staff can bedetrimental to a prison’s operations. On the other hand,

having relatively unstressed, involved, satisfied, and com-

mitted staff is a desired outcome. To reach this outcome,

researchers and administrators need information on the

factors that help shape the job stress, job involvement,

job satisfaction, and organizational commitment of prison

staff.

Working in a prison differs from working in other

types of organizations. The prison work environment is

unique because it involves confining individuals against


0362-3319/© 2015 Western Social Science Association. Published by Elsevier Inc. All rights reserved.

http://localhost/var/www/apps/conversion/tmp/scratch_4/dx.doi.org/10.1016/j.soscij.2015.10.001http://localhost/var/www/apps/conversion/tmp/scratch_4/dx.doi.org/10.1016/j.soscij.2015.10.001http://www.sciencedirect.com/science/journal/03623319http://www.elsevier.com/locate/soscijmailto:[email protected]:[email protected]:[email protected]:[email protected]://localhost/var/www/apps/conversion/tmp/scratch_4/dx.doi.org/10.1016/j.soscij.2015.10.001http://localhost/var/www/apps/conversion/tmp/scratch_4/dx.doi.org/10.1016/j.soscij.2015.10.001mailto:[email protected]:[email protected]:[email protected]:[email protected]://www.elsevier.com/locate/soscijhttp://www.sciencedirect.com/science/journal/03623319http://localhost/var/www/apps/conversion/tmp/scratch_4/dx.doi.org/10.1016/j.soscij.2015.10.001http://localhost/var/www/apps/conversion/tmp/scratch_4/dx.doi.org/10.1016/j.soscij.2015.10.001


2/11






2 E.G. Lambert et al. / The Social Science Journal xxx (2015) xxx–xxx

their will—individuals who have been convicted of violat-

ingcriminallaw. As Armstrong and Griffin (2004) point out,

“Few other organizations are charged with the central task

of supervising and securing an unwilling and potentially

violent population” (p. 577). As the work environment in

prisons varies from that found in most other organizations,

the factors which influence job stress, job involvement, job

satisfaction, and organizational commitment may alsovary

compared to other organizations. Research needs to iden-

tify factors that mitigate job stress and contribute to the

prison staff’s job involvement, job satisfaction, and organi-

zational commitment.

The literature suggests that social support is important

for staff to deal with the unique strains and challenges

encountered in a prison work environment (Keinan &

Maslach-Pines, 2007; Whitehead & Lindquist, 1986). Social

support should help staff deal with strains that would

otherwise increase job stress and decrease job involve-

ment, job satisfaction, and organizational commitment.

There are different types of social support, and the liter-

ature is limited and unclear about what types of social

support are linked with job stress, job involvement, job

satisfaction,and organizationalcommitment amongprison

staff. This study’s objective is to explore the nature of the

relationship between administrative support, supervisory

support, coworker support, and family and friends support

with job stress, job involvement, job satisfaction, and orga-

nizational commitment. Expanding knowledge of these

relationships provides a framework for the future devel-

opment of research in this area and provides correctional

administrators with information on how different forms of

social support affect prison staff.

1. Literature review

1.1. Job stress

Accordingto Mattesonand Ivancevich (1987), “thereare

literallyhundredsof definitions for stress to be found in the

research and professional literature. Virtually all of them

can be placed into one of two categories, however: stress

can be defined as either a stimulus or a response” (p. 10).

Stressors arenegativestimuli that cause strainfor a person,

which ultimately can result in stress (Cullen, Link, Wolfe,

& Frank, 1985). Job stress is psychological strain leading to

leading to job-related hardness, tension, anxiety, frustra-

tion, and worry arising from work (Misis, Kim, Cheeseman,

Hogan, & Lambert, 2013). Job stress can be harmful over

time, leading to increased mental withdrawal from the job,

reduced interactions with clients and coworkers, increased

conflict with family and friends, absenteeism, substance

abuse, turnover, burnout, health/medical problems, and

even premature death (Cheek, 1984; Cheek & Miller, 1983;

Lambert, Edwards, Camp, & Saylor, 2005a; Matteson &

Ivancevich, 1987; Mitchell, MacKenzie, Styve, & Gover,

2000; Slate & Vogel, 1997; Whitehead, 1989; Whitehead

& Lindquist, 1986; Woodruff, 1993). Job stress is harm-

ful to both staff and the prison organization, and its

antecedents need to be studied. Social support may be one

such antecedent.

1.2. Job involvement

Job involvement is the level of psychological identifica-

tion with a job (Kanungo, 1982a,b). As Elloy, Everett, and

Flynn (1992) note, it refers to a cognitive bond people can

form with their jobs. Paullay, Alliger, and Stone-Romero

(1994) point out that job involvement is a psychological

state wherein an individual “is cognitively preoccupied

with, engaged in, and concerned with one’s present job”

(p. 224).DeCarufeland Schaan (1990) notethat“an individ-

ual with a high degree of job involvement would place the

job at the center of his/her life’s interests. The well-known

phrase ‘I live, eat, and breathe my job’ would describe

someone whose job involvement is very high” (p. 86). Per-

sons with low job involvement focus on interests other

than their work (Hogan, Lambert, & Griffin, 2013). Elloy

et al. (1992) contend that job involvement is a measure

worklife’s quality.ChenandChiu(2009)point outthatpeo-

ple with “high job involvement are more independent and

self-confident–they not only conduct their work in accor-

dance with the job duties required by the company but are

also more likely to do their work in accordance with the

employees’ perception of their own performance” (p. 478).

Pfeffer (1994) contends that job involvement translates to

organizational effectiveness in the long run. Additionally,

job involvement is linked with reduced turnover intent

among jail staff (Lambert & Paoline, 2010). There is a need

to examine possible antecedents of prison staff job involve-

ment, and social support may be an important antecedent.

1.3. Job satisfaction

Locke (1976) defines job satisfaction as “a pleasurable

or positive emotional state resulting from the appraisal

of one’s job or job experiences” (p. 1300). Weiss (2002)

contends job satisfaction is the summative positive and

negativeemotions arising fromthe job. Simply, job satisfac-

tion is “the extent to which people liketheir jobs” (Spector,

1996, p. 214). In all these definitions, job satisfaction is

an affective/emotional response by an employee concern-

ing his/her particular job and whether the employee likes

the job. Job satisfaction is a salient and powerful work-

place concept. Low levels have been found to be associated

with absenteeism, turnover, and job burnout among cor-

rectional staff (Byrd, Cochran, Silverman, & Blount, 2000;

Whitehead & Lindquist, 1986; Wright, 1993). Conversely,

high levels are related to improved work performance,

organizationalinnovation,greatercompliance withorgani-

zational rules, greater support for rehabilitationof inmates,

and greater life satisfaction among correctional staff (Fox,

1982; Lambert, 2004; Lambert & Hogan, 2010; Lambert

et al., 2009; Lambert, Hogan, Paoline, & Baker, 2005b;

Robinson, Porporino, & Simourd, 1992). Given its signif-

icance, it is important to explore how different forms of

social support may be associated with job satisfaction.

1.4. Organizational commitment

Organizational commitment is the bond between an

employee and the employing organization (Mowday,

Porter, & Steers, 1982). Continuance and affective

http://localhost/var/www/apps/conversion/tmp/scratch_4/dx.doi.org/10.1016/j.soscij.2015.10.001http://localhost/var/www/apps/conversion/tmp/scratch_4/dx.doi.org/10.1016/j.soscij.2015.10.001


3/11






E.G. Lambert et al. / The Social Science Journal xxx (2015)xxx–xxx 3

commitment are two major forms of commitment, and

they differ in how the bond is formed (Hogan et al., 2013).

Continuance commitment refers to a bond that occurs due

to investments made in the employing organization, such

as social relationships, pension, salary and benefits, and

nontransferable job skills (Allen & Meyer, 1990). These

investmentsbond theperson to theorganization andresult

in a desire to stay with the organization (Garland, Hogan,

Kelley, Kim, & Lambert, 2013). With affective commitment,

a voluntary bond is formed by positive work experiences

(Allen & Meyer, 1990). The concept encompasses the ele-

ments of loyalty to theorganization, identification with the

organization, such as pride in the organization and inter-

nalization of organizational goals, and involvement in the

organization, such as personal effort made for the sake

of the organization (Hogan et al., 2013; Mowday et al.,

1982). With continuance commitment, a worker bonds

with the organization because they must do so. With affec-

tive commitment, a worker bonds with the organization

because they choose to do so (Meyer & Allen, 1991). Of the

two, affective commitment tends to have far more salient

outcomes for both employees and organizations. Further,

continuance commitment has been found to be associ-

ated with some negative consequences, as burned out

employees may feel trapped in the job and view quitting

as an attractive but untenable option. As such, the current

study focused on affective commitment, a concept gener-

ally viewed as a highly desirable outcome (Lambert, Kim,

Kelley, & Hogan, 2013). Among correctional staff, affective

commitment has been inversely linked with absenteeism

and turnover and positively associated with job perfor-

mance, life satisfaction, and organizational citizenship

behaviors, which is going beyond what is expected at work

(Camp, 1994; Culliver, Sigler, & McNeely, 1991; Lambert

et al., 2005a, 2013; Lambert, Hogan, & Griffin, 2008).

1.5. Social support

Social support refers to a network of connections with

other human beings that can provide assistance, support,

andhelpforaperson(Cohen, Underwood, & Gottlieb, 2000;

Harvey, 2014; Lambert, Altheimer, & Hogan, 2010). A lack

of social support can lead employees to feel isolated and

alone at work, exacerbatingthe effects of workplace stress-

ors (Ileffe & Steed, 2000). In addition, social support can

provide resources for people to deal more effectively with

stressors,ultimately reducing job stress (Neveu, 2007). Fur-

thermore, social support is a valuable resource because it

provides psychological support, assistance, feedback, and

motivation for employees (Lambert et al., 2010). Most peo-

pleare social creatures, and social support can be a positive

element for employees (Cohen et al., 2000), but its benefits

extend to both employees and employers. Social support

systems can allow innovation to occur, which can result

in quicker solutions to workplace issues, allowing work to

become more productive and enjoyable. Conversely, a lack

of socialsupportcan be seen as a form of resource depletion

for prison workers, detractingfrom positive outcomes from

work. Social support can be an excellent way to boost self-

confidence and self-esteem (Lambert et al., 2010). Social

support can increase the likelihood that a staff member

will bond with thejob. In addition,the positive experiences

with support can result in greater satisfaction from the job.

Based on the social exchangetheory, affectivecommitment

arises due to positive work experiences that allow individ-

uals to see the organization in a favorable light (Colquitt

et al., 2013).

There is a gap in the literature in terms of which

forms of social support are potential antecedents of prison

staff’s job stress, job involvement, job satisfaction, and

organizational commitment. Social support comes in dif-

ferent forms (Lee & Ashford, 1996; Whitehead & Lindquist,

1986). The social support’s major forms are administra-

tive support, supervisory support, coworker support, and

family andfriends support. Support fromcoworkers, super-

visors, and management represents intra-organizational

social support systems, while support from family and

friends represents an extra-organizational support system

(Lambert et al., 2010; Ray & Miller, 1994).

Administrative support refers to the staff’s percep-

tion that they are supported by the prison’s manage-

ment/administration (Garland & McCarty, 2006; Griffin,

2006). Administrative support can help staff deal with

stressors and being stressed from work. A lack of admin-

istrative support can make work difficult and straining.

Cheek and Miller (1983) contend that a lack of adminis-

trative social support can lead to staff feeling pressured

by both management and the inmates, resulting in being

between the proverbial rock and a hard place. They term

this situation a “double-bind” of stress for correctional

staff. Administrative support sends a message to staff that

they are valued and respected by organization. In addition,

administrative support can help staff be more successful

in their jobs, possibility resulting in greater identification

with the job and satisfaction. Administrative support, or

a lack thereof, for most staff will represent the organiza-

tion. Eisenberger, Huntington, Hutchison, and Sowa (1986)

report that employees “personify the organization, viewing

actions by agents of the organization as actions of theorga-

nization itself” (p. 504). In a sense, administrative support

reflects the level of commitment an organization has for

its employees. If staff perceive that theadministration sup-

ports them, it is easier for them to form a cognitive bond

with the organization. These postulations have a basis in

the literature. Administrative support is inversely associ-

ated with correctional staff job stress (Armstrong & Griffin,

2004; Auerbach, Quick, & Pegg, 2003; Griffin, 2006). Like-

wise, administrative support is positively associated with

job satisfaction and organizational commitment (Lambert

& Hogan, 2009).

Supervisory support focuses on staff perceptions of the

level that their supervisor supports them (Cullen et al.,

1985). Supervisory support can help remove stressors

or reduce their negative effects, which should translate

to lower stress from the job (Brough & Williams, 2007).

Supervisors can help staff see challenges in a more positive

light. Lee and Ashford (1996) note that “with the right kind

of supervisory support, workers may come to perceive

ambiguous role expectations as opportunities to carry

out their own initiatives (potential gains) rather than as

restrictions on their actions (certain losses)” (p. 131). On

the other hand, a lack of supervisory support can be a



4/11







straining and frustrating experience, which over time can

lead to job stress (Lambert & Hogan, 2009). Supervisory

support is important for employees to succeed in their

jobs. It is a resource which can help staff do their jobs in

an effective manner. Being productive tends to provide

people with a sense of pride and accomplishment. In the

end, prison staff are probably more likely to feel that the

job is meeting some of their needs and wants, and, as

such, will view the job in a more favorable light, increasing

the level of job involvement and satisfaction. Supervisors

are also seen as the prison organization’s representatives

(Lambert, 2004). Supervisory support sends a message

that staff are valued; thus, supervisory support likely

will increase favorable views of the organization, and, in

turn, strengthen the bond to it. On the contrary, a lack of

supervisory support can be a type of resource withdrawal

that leads to staff to question the organization, and, in

the long run, drive a wedge between staff and the prison

organization. Supervisory support is associated with lower

job stress and increased job satisfaction and organizational

commitment of correctional staff (Cullen et al., 1985;

Grossi, Keil, & Vito, 1996; Lambert, 2004; Lambert &

Hogan, 2009; Van Voorhis, Cullen, Link, & Wolfe, 1991).

Coworker support refers to the perceptions of support

from peers at work (Liaw, Chi, & Chuang, 2010). Coworker

support can help deal with stressors and facilitate man-

agement of minor problems quickly before they get out

of hand (Bakker & Demerouti Schaufeli, 2005). A lack of

peer support can be a trying experience in itself, raising

the degree of frustration encountered at work. Coworker

support should help reduce job stress among prison staff.

Support from work peers is also a resource for many peo-

ple (Lambert et al., 2010). It allows a chance to discuss

work issues and, at times, to “vent.” Cherniss (1980) points

out that “when one can discuss work experiences with

colleagues, those experiences often become more inter-

esting and meaningful” (p. 120). It provides a place for

advice and social companionship, as well as a support sys-

tem (Whitehead & Lindquist, 1986). In the end, coworker

support can make thejob enrichingand enjoyable, increas-

ing the level of job involvement and job satisfaction for

staff. A lack of support for coworkers could lead to less

identification with the job, as well as having unmet needs

from the job. Positive experiences from coworker support

may spillover and allow staff to seethe prison organization

in a more positive light, promoting greater attachment to

the organization. Having positive relations with coworkers

is associated with increased satisfaction and lowered job

stress among jail staff (Paoline, Lambert, & Hogan, 2006),

and lower job stress among Australian prison staff (Dollard

& Winefield, 1998).

Family and friends social support deals with the per-

ception by staff that their non-work significant others

care and help them with work matters (Lambert et al.,

2010). Family and friends are the primary social support

for people outside the workplace (Adams, King, & King,

1996). This form of support can help individuals deal

with work problems (Kurtz, 2008). Working in a prison

can be a trying experience (Armstrong & Griffin, 2004).

Having support at home may help reduce stress from

work. This form of support may also provide a safe haven

for staff to escape the trying work experiences, even if it is

a temporary respite, and, in turn, may help alleviate stress

from the job. In addition, having support from family and

friends for working in a prison may give a person a sense

of pride, allowing them to have a greater bond with the job

and the organization. In addition, having positive social

support from family and friends concerning the job may

result in positive feelings about their lives and their jobs,

spilling over to raise job satisfaction levels (Kwok, Cheng,

& Wong, 2014). Finally, family and friends support can

allow prison staff to lead a more balanced life, which, in

the end could result in general positive feelings that help

raise the levels of job involvement, job satisfaction, and

affective commitment (Cordes & Dougherty, 1993).

1.6. Research focus

While limited research exists on the effects of the dif-

ferent forms of social support on correctional staff, gaps

remain to be addressed. First, there has not been sufficient

research to date to uphold derive firm conclusions. Sec-ond, there has been far more research on the effects of

supervisory support than there has been for administra-

tion, coworker,and friends and family support. There needs

to be more research on all four forms of social support

among prison staff in order to understand the role social

support plays. Third, no published study has examined the

relationships of the four forms of support with job stress,

job involvement, job satisfaction, and organizational com-

mitment among prison staff in the same study. As a result,

theknowledge about theeffects of socialsupport in limited

and fragmentary. For example, one form of support may

be more important in shaping a particular outcome area

among prison staff than the other forms of support. Thisstudy’s purpose is to address these gaps in knowledge.

2. Method

2.1. Participants

All the available staff at a private Midwestern prison for

juvenile offenders were provided a survey packet, which

contained a cover letter, survey, and return envelope. The

cover letter explained the study’s purpose, that participa-

tion was voluntary, how to participate in the study, how to

returnthe surveys,and stressedthatall responseswould be

anonymous and would be kept confidential. At the time of the survey, there were 220 employees assigned to work at

the prison, but due to leave, such as sick leave and admin-

istrative leave, only 200 were available to be provided the

survey packet. A total of 160 usable surveys were returned,

which represents a response rate of 80% of those given the

packet and 72.7% of those who worked at the prison1. The

1 The survey measured a wide array of workplace perceptions, views,

attitudes, intentions, and behaviors of employees. Because of the wealth

of information from the survey, other studies have been conducted using

differentparts of the survey. The full citations of thesestudiesare available

upon request. None of these previous studies examined the impact of the

fourforms of social support on jobstress,job involvement,job satisfaction,and organizational commitment together.



5/11







prison housed approximately 450 juvenile offenders under

the age of 20 who were juveniles sentenced in adult court.

The prison was a high security closed facility.

Employees came from all positions except administra-

tion. Demographic characteristic of the respondents are

listed in Table 1. Institutional records indicate that, at

the time of the survey, the demographic breakdown of

the prison’s staff was approximately 81% White and 61%

male; the staff’s median age was about 35, the median

tenure was about 20.64 months, and about two-thirds of

theemployees held a custody position.As such, therespon-

dents appear to be demographically representative of the

staff at the private correctional facility.

2.2. Variables

2.2.1. Dependent variables

The dependent variables are additive indexes mea-

suring job stress, job involvement, job satisfaction, and

organizational commitment. All the items were answered

using a five-point Likert response scale ranging from

strongly disagree (coded 1) to strongly agree (coded 5).

The responses to the items were summed together to form

indexes, and Cronbach’s alpha values were calculated and

are listed in Table 1. The job stress index comprises five

items from Crank, Regoli, Hewitt, and Culbertson (1995).

For example, two job stress items are “I am usually under a

lot of pressure when I am at work” and “When I’m at work

I often feel tense or uptight.” The job involvement index

comprises three items adapted from Kanungo (1982a,b).

For example, two job involvement items are “I live, eat,

and breathe my job” and “The most important things that

happento me inmy life occur at work.” Five items, adapted

from Brayfield and Rothe (1951), are used to measure job

satisfaction. For example, two job satisfaction items are “I

like my job better than the average worker does” and “Most

days I am enthusiastic about my job.” The affective orga-

nizational commitment index comprises six items from

Mowdayet al.(1982). For example,two commitment items

are “I feel very little loyalty to this prison” (reverse coded)

and “I find that my values and the prison’s values are very

similar.”

2.2.2. Independent variables

Fourtypes of support are measured:administrative sup-

port, supervisory support, coworker support, and family

and friends support. The support items were answered

using a Likert scale ranging from strongly disagree (coded

1) to strongly agree (coded 5), and the responses to the

items were summed together to form indexes. See Table 1

for Cronbach’s alpha values. The administrative support

index comprises two items. For example, one item is “For

the most part, management at this prison supports its

workers.”The index for supervisory support comprisesfour

items. For example, three supervisory support items are

“Supervisors at this prison are supportive of employees,”

“At this facility, supervisors often criticize employees over

minor things” (reverse coded), and “My supervisor looks

out for my personal welfare.” The coworker support index

comprises six items. For example, three of the items are

“I am able to discuss problems with my coworkers,” “My

coworkers provide me support in solving personal prob-

lems”, and “The people I work with are friendly.” An index

for support for family and friends comprises two items. For

example, one item is “When my job gets me down, I know

that I can turn to family and friends for support.” Finally,

the personal characteristics of gender, age, tenure, posi-

tion, educational level, race, and supervisory status were

included in the study more as control than explanatory

variables. See Table1 for how these personalcharacteristics

are coded.

3. Results

The descriptive statistics for the variables in this study

are presented in Table 1. There appears to be significant

variationin both the dependent and independent variables.

None of the variables are constants. The median and mean

are similar to one another for the variables, suggesting

that the variables were normally distributed. Cronbach’s

alpha, a measure of internal reliability, was above .70 for all

the indexes, which indicates acceptable internal reliability

(Gronlund, 1981). The items used to measure job stress, job

involvement, job satisfaction, and organizational commit-

mentwere enteredintoan exploratory factor analysisusing

generalized least squares method with an equamax rota-

tion. The items for each latent concept loaded upon same

factorwithfactor loading scoresof .41or higher, andmostly

.50 or higher. For example, thethreejob involvement items

loaded on a single factor with factor loading scores of .61,

.84, and .83. The four factors explained approximately 64%

of the variance of the items. Similarly the 14 social sup-

port items were entered into an exploratory factor analysis

using generalized least squares with an equamax rotation.

Four factors explained for about 55% of the variance. The

items for each latent concept load upon same factor with

factor loading scores of .41 or higher, and mostly .50 or

higher. For example, the two administrative support items

loaded on one factor, with .67 and .61 factor loading scores.

The Pearson’s correlation coefficient results are pre-

sented in Table 2. Tenure, administrative support, super-

visory support, and coworker support have statistically

significant correlations with job stress. Increases in tenure

are associated with higher levels of stress from the job.

Conversely, increases in administrative, supervisory, and

coworker support are related to lower reported stress from

the job. Gender, age, position, educational level, race, and

family and friends support each have non-significant cor-

relations with job stress.

Position, supervisory status, administrative support,

supervisory support, and coworker support have sig-

nificant correlations with job involvement. Correctional

officers report, on average, lower levels of involvement

with thejob as compared to non-custody staff. Supervisors,

in general, report higher job involvement than their non-

supervisory counterparts. Administrative, supervisory, and

coworker support have positive correlations, which means

an increase in each type of social support is linked with

a rise in job involvement. Gender, age, tenure, educa-

tional level, race, and family and friends support have

non-significant correlations.



6/11







Table 1

Descriptive statistics for study variables.

Variable Description Md Min Max Mean SD

Gender 41% women (coded 0) 1 0 1 .59 .49

59% men (coded 1)

Age Measured in continuous years 33 19 68 35.77 10.82

Tenure Tenure at the prison in months 17 1 53 20.64 13.84

Position 38% non-CO (coded 0) 1 0 1 .62 .49

62% CO (coded1)

Educ level 53% have no college degree (coded 0) 0 0 1 .47 .50

47% have a college degree (coded 1)

Race 21% Nonwhite (coded 0) 1 0 1 .79 .50

79% White (coded 1)

Supervisor 79% non-supervisors (coded 0) 0 0 1 .21 .41

21% supervisors (coded1)

Admin support 2 item additive index, ˛= .84 6 2 10 5.31 2.21

Super support 4 item additive index, ˛= .80 11 4 20 10.92 3.67

Cowrk support 6 item additive index, ˛= .81 20 7 27 19.52 4.53

Family support 2 item additive index, ˛= .70 8 2 10 7.12 1.72

Job stress 5 item additive index, ˛= .82 14 6 25 14.29 4.51

Job involve 3 item additive index, ˛= .81 6 3 15 5.90 2.51

Job sat 5 item additive index, ˛= .92 19 5 25 18.34 4.67

Org commit 6 item additive index, ˛= .88 19 10 26 18.14 4.00

Note. Md stands for median value, Min stands for minimum value, Max stands for maximum value, SD stands for standard deviation, CO stands for

correctional officer, Educ stands for educational, Supervisor for supervisory status, Admin stands for administrative, Super stands for supervisory, Cowrkstands for coworker, Family stands for family and friends, involve stands for involvement, sat stands for satisfaction, Org commit stands for organizational

commitment, and stands for Cronbach’s alpha, a measure of internal reliability. For Educ level, having a college degree includes those who had earned

an Associate’s/Vocational degree, a Bachelor’s degree, and Graduate/Professional degree. The data for this table are from a survey of 160 staff members at

private Midwestern prison for juvenile offenders tried as adults.

Table 2

Pearson correlation matrix for study variables.

Variables 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14.

1. Gender 1.00

2. Age −.17* 1.00

3. Tenure .00 .07 1.00

4. Position .28** −.53** −.20** 1.00

5. Educ lev −.13 .14 .05 −.29** 1.00

6. Race .08 −.05 .06 −.02 .05 1.007. Supervisor .12 .16* .39** −.47** .16* .04 1.00

8. Adm sup .02 .29** −.09 −.32** .09 .11 .29** 1.00

9. Sup −.05 .21** −.05 −.30** .15 .06 .27** .76** 1.00

10. Co sup .06 .13 −.13 −.26** .05 .11 .02 .49** .42** 1.00

11. Fam sup .06 −.05 −.04 −.01 −.08 −.14 .03 .20* .13 .35** 1.00

12. Job str −.12 −.02 .20* .07 .01 .04 −.10 −.49** −.51** −.31** −.11 1.00

13. Job inv .03 −.01 .12 −.27** .13 .05 .26** .37** .43** .21** −.10 −.34** 1.00

14. Job sat .07 .18* −.07 −.30** .18* .01 .28** .57** .57** .40** .09 −.70** .41** 1.00

15. Org com .05 .17* −.04 −.24** .12 .00 .23** .66** .68** .46** .16* −.53** .46** .75**

Note. See Table 1 for a description of the variables and how they were coded. Educ lev stands for educational level, Supervisor for supervisory status, Adm

sup stands for administrative support, Sup for supervisory support, Co sup for coworker support, Fam sup for family and friends support, Job str for job

stress, Job inv for job involvement, Job sat for job satisfaction, and Org com for organizational commitment. The data for this table are from a survey of 160

staff members at private Midwestern prison for juvenile offenders tried as adults.* p≤ .05.

** p≤ .01.



7/11







Table 3

Multivariate ordinary least squares regression results.

Variables Job stress Job involvement Job satisfaction Org commitment

B ˇ B ˇ B ˇ B ˇ

Gender −1.00 −0.11 0.41 0.08 .96 .10 .43 .05

Age 0.05 0.11 −0.06 −.26** −.01 −.03 −.01 −.02

Tenure 0.04 0.14 0.02 0.12 −.03 −.08 .01 .02

Position 0.14 0.02 −1.24 −.24

*

−.86 −.09 .00 .00Educ level 0.57 0.06 0.08 0.02 .77 .08 .11 .01

Race 1.13 0.1 −0.37 −0.06 −.87 −.08 −.89 −.09

Supervisor −0.11 −0.01 0.07 0.01 1.09 .10 .16 .02

Admin support −0.48 −.24* 0.17 0.15 .47 .22* .52 .28**

Super support −0.41 −.34** 0.21 .30** .37 .29** .44 .40**

Cowrk support −0.07 −0.07 0.04 0.07 .16 .16* .16 .18*

Family support 0.11 0.04 −0.31 −.21** −.17 −.06 −.07 −.03

R-Squared .35** .31** .42** .54**

Note. See Table 1 f or a description of thevariablesand howthey were coded.B stands for the unstandardized regression coefficient,ˇ for the standardized

regression coefficient, Educ for educational,Supervisorfor supervisory status, Admin for administrative,Super for supervisory, Cowrk for coworker, Family

for family and friends, and Org for organizational. The data for this table is from a survey of 160 staff members at private Midwestern prison for juvenile

offenders tried as adults.* p≤ .05.

** p≤ .01.

Age, position, educational level, supervisory status,

administrativesupport,supervisory support, and coworker

support have significant correlations with job satisfaction.

Older workers tend to report higher job satisfaction than

younger workers. Correctional officers tend to report lower

levels of job satisfaction compared to non-custody staff.

More highly educated and supervisory staff report higher

levels of job satisfaction compared to their less educated

and non-supervisory coworkers. Administrative support,

supervisory support, and coworker support have positive

correlations with job satisfaction, meaning an increase in

each type of social support is linked with a rise in job satis-

faction. Allthe other variables, includingfamily and friends

support, have non-significant correlations with job satis-

faction.

Age, position, supervisory status, administrative sup-

port, supervisory support, coworker support, and family

and friends support have positive correlations with

commitment. Correctional officers, in general, report

lower commitment than their non-custody counterparts.

Supervisors on average have higher levels of affectivecom-

mitment than non-supervisory staff. Increases in age and

the four forms of support are associated with greater levels

of commitment to the organization.

Along with the personal characteristics, the social sup-

port variables were entered as independent variables into

Ordinary Least Squares (OLS) regressionequations with job

stress, job involvement, job satisfaction, and organizational

commitment as the dependent variables. The results for

the four OLS regression equations are presented in Table 3.

While not reported, the tolerance and variance inflation

factor statistics indicate no problem of multicollinearity

among the independent variables. High multicollinearity

occurs when an independent variable shares a very large

part of its variance with the other independent variables in

the regression equation, and this impedes the estimation

of the effects of the affected independent variables on the

dependent variable (Berry, 1993).

For the job stress equation, the R-Squared statistic is

.35, which means that the independent variables explained

about 35% of the variance observed in the job stress vari-

able. Administrative support and supervisory support have

significant relationships with job stress, and direction of

the relationship is negative for both. In other words, an

increase in either form of support is associated with lower

reported stress from the job. The regression coefficients

in the columns in Table 3 estimates the magnitude of the effect of an independent variable on the dependent

variable. For the job stress equation, supervisory sup-

port has the greatest effect, followed by administrative

support.

About 31% of the variance in the job involvement

variable is accounted for by the independent variables.

Age, position, supervisory support, and family and friends

support each have significant relationships with job

involvement. Correctional officers generally report lower

job involvement. Increases in age and support from fam-

ily and friends are linked to lower reported involvement

in the job, while greater supervisory support is linked

to higher involvement. Among the significant variables,

supervisory support has the largest sized effect, followed

closely by age, position, and family and friends sup-

port.

Approximately 42% of the variance in the job sat-

isfaction index is accounted for in the OLS regression

equation.Administrative support, supervisory support,and

coworker support have significant associations with the

dependent variable, and the direction of the relation-

ship for each is positive. Among the significant variables,

supervisory support has the largest effect, followed by

administrative support and then coworker support.

Approximately 54% of the observedvariancein the orga-

nizational commitment variable is accounted for in the

multivariateequation. Administrativesupport, supervisory

support, and coworker support have a significant posi-

tive relationship with commitment to the organization.



8/11







Similar to the job satisfaction results, supervisory support

has the largest sized effect, followed by administrativesup-

port, and then coworker support.

4. Discussion and conclusion

As suggested by limited past research (Keinan &

Maslach-Pines, 2007; Lambert & Hogan, 2009; Whitehead

& Lindquist, 1986), data from this study indicate that

perceived social support is an important antecedent of

such favorable outcomes among correctional staff; how-

ever, the extent to which different types of social support

are related to the outcomes of job stress, job involvement,

job satisfaction, and organizational commitment–variables

known to be integral to agency operations and effective-

nessin corrections–was unknown until this study (Lambert

et al., 2005a, 2008; Pfeffer, 1994; Wright, 1993). Correc-

tional administrators seeking to improve these outcomes

among staff can exercise more direct control over some

forms of support, such as administrative and supervisory,

than other forms, such as coworkers and family/friends.

The existing literature is unclear about how administra-

tive, coworker, and family/friends support relates to job

outcomes; most past research has concentrated on super-

visory support. Finally, previous research on social support

hasbeen fragmentary, in that no singlestudy hasexamined

the construct in relationship to job stress, job involve-

ment, job satisfaction, and organizational commitment. In

short,knowledgeis very limited about whichtypes of social

support influence which job outcomes among correctional

staff.

Findings from the present study shed light on these

issues. First, social support is far more important in shap-

ingthe four job outcomes than thepersonal characteristics.

This is consistent with what research on correctional

staff has generally established about the importance of

work environment factors relative to personal factors with

regards to job attitudes and behaviors (Matz, Wells, Minor,

& Angel, 2013; Minor, Wells, Lambert, & Keller, 2014). Only

two of the personal variables studied here are significantly

related to a single job outcome, job involvement. Older

employees and those in correctional officer positions dis-

played lower job involvement. A plausible interpretation

of the age effect is that older correctional staff may be

more likely than younger ones to have developed capac-

ity to keep work in perspective relative to other priorities,

to have devised routines and strategies for getting work

done efficiently, and also to have developed alternative

interests and avenues of psychological identification and

for deriving life satisfaction, such as activities of their chil-

dren, hobbies, and so forth. By contrast, younger staff who

are newer to the world of full-time work may experience

their jobs, at least initially, as more engulfing. They may

alsobe relatively more dependent than their oldercounter-

parts on their jobs for money due to having had less time

to establish savings, and some might see engrossment in

work as a pathway to loftier career ambitions. Compared

with other jobs in prison, correctional officer work is usu-

ally more regimented with clearly demarcated times for

starting and ending shifts; thus, correctional officers could

be more likely to see their work as an occupation instead

of a profession, and this could help account for the lower

job involvement among correctional officers found in this

study. These, of course, are an untested postulations.

Job involvement’s other two significant correlates are

supervisory support and family support. Staff members

who feel their supervisor supports them are likely to iden-

tify more closely with their jobs and are less likely to feel

alienated than those who feel unsupported; these staff are

probably more likely to experience motivation to become

immersed in their work. Research indicates that a moti-

vating work climate, which is something that a supportive

supervisor can inspire, promotes job involvement (Brown

& Leigh, 1996).

By contrast, the data from this study show an inverse

relationship between support from family/friends and job

involvement. People who experience less support from

their family and friends may gravitate more closely toward

identifying with their jobs. Additionally, Lambert (2008)

indicates that greater levels of job involvement are linked

to a greater probability of work-induced family conflict

among correctional staff. That is, staff who feel overloaded

by the job role experience greater work-on-family conflict

(Lambert, Minor, Wells, Lambert, & Hogan, 2015). Social

support from familyand friends could function as a double-

edged sword with respect to job involvement. On the one

hand, such support could help the employee maintain a

proper perspective on work, causing the employee to bal-

ance work with other life domains and refrain from an

unhealthy obsession with his/her job. On the other hand,

relations with family and friends could lead the employee

to become overly distracted from work responsibilities by

outside interests and commitments (Hogan et al., 2013).

Further research is needed to improve understanding of

the relationship between family/friends support and job

involvement.

Consistent with past research (Brough & Williams,

2007; Cheek & Miller, 1983; Cullen et al., 1985; Griffin,

2006; Keinan & Maslach-Pines, 2007), correctional staff

in the present study who feel more supported by their

administrationand supervisors report lower job stress than

those whoperceive less support from these sources. As line

staff members’ bureaucratic superiors, administrators and

supervisors wield considerable power and influence, and

feeling that these persons are unsupportive can make the

job more stressful for staff. A lack of support from adminis-

trators and supervisors is likely to make staff feel devalued

and vulnerable, to exacerbate workplace stressors, and to

deprivestaff of a valuableresourcefor coping withstressors

(Ileffe & Steed, 2000; Lambert et al., 2010; Neveu, 2007).

Similarly, this study’s findings show that higher lev-

els of social support from administrators, supervisors, and

coworkers were significant antecedents of both job satis-

faction and organizational commitment. These results are

consistent with exchange theory principles (Colquitt et al.,

2013) as well as with findings from past studies (Lambert &

Hogan, 2009; Paoline et al., 2006). Feeling supported on the

job by bureaucraticsuperiors and peers can lead employees

to reciprocate by liking their jobs and developing loyalty

and solidarity with the organization. By contrast, perceiv-

ing these groups as non-supportive probably renders the

job unsatisfying and impairs bonding to the organization.



9/11







Thefindings of thepresentstudyare consistent with the

conclusion that administrative, supervisory, and coworker

support operate in tandem, with supervisory support

exerting the strongest and most consistent effects across

the four outcomes studied, followed by administrative

support, which is related to all outcomes except job

involvement, and finally support of coworkers, which is

related to job satisfaction and organizational commitment

only. As an extra-organizational source of social support

(Lambert et al., 2010; Ray & Miller, 1994), family/friends

support is not significantly related with the outcomes of

job stress, job satisfaction, or organizational commitment;

it is only related to job involvement. This pattern of find-

ings underscores the importance of differentiating forms of

social support and examining these forms simultaneously,

as not all types relate to the same job outcomes.

Moreover, it seems that intra-organizational sources

of support can be differentiated according to the posi-

tion of the parties providing, or not providing, support

to a staff member. Such differentiation can be conceptu-

alized as having vertical and lateral dimensions. On the

vertical dimension is support coming from those in the

organization positioned above the staff member, includ-

ing supervisors and higher level administrators, as well

as those positioned below the staff member. The lateral

dimension involves support from peer workers. Because

parties positioned above a staff member can be a pressing

source of job demands, such as daily expectations, mon-

itoring and evaluation of performance, and so forth, and

also a resource for coping with job strains, it makes sense

that low levels of perceived support from administrators

and supervisors might promote more job stress than a lack

of lateral support from coworkers. Similarly, support com-

ing from immediately above an employee in the vertical

chain in terms of supervisory support would logically have

a greater likelihood of leading to high job involvement

than support coming from higher, further removed levels

in terms of administrative support, or support from lateral

levels in terms of coworker support. Also, the present find-

ings suggest that both the vertical and lateral dimensions

of intra-organizational support are important for fostering

a liking for the job and commitment to the organization

among correctional staff. Although the supervisor variable

is notsignificantly related to theoutcomes examinedin this

study is indicated in Table 3, f uture research should attend

to how perceived support from subordinates is related to

these and other job outcomes.

The current study has limitations. This was a single

study of staff at a private prison. The vast majority of pris-

onsin theU.S. arepublic facilities (Lambert & Hogan,2009).

Results could differ among staff at a public prison. More-

over, the private prison held juveniles sentenced as adults,

which is a uniqueinmatepopulation notfoundat many pri-

vate or even public prisons. Dealing with juvenile offenders

is a differentfocusthan dealing with adult offenders. Treat-

ment was a major focus at this private prison as state law

requires educational services be provided to all inmates

under 18 years of age. The unique inmate population

and treatment efforts could have affected the associations

between the different forms of social support and the out-

come areas of job stress, job involvement, job satisfaction,

and organizational commitment. Research at other prisons

and types of correctional facilities, such as jails, public pris-

ons, adult facilities, and so forth, is needed to determine

whether the findings can be replicated. With replication,

the effects of different forms of social support on prison

staff can be understood more fully. Additionally, it remains

to be determined if effects are contextual and situational,

varying across different types of correctional facilities. The

current study uses a cross-sectional design of collectingthe

data at one point in time, which precludes demonstrat-

ing causal relationships. Longitudinal designs are needed

to demonstrate the casual process of the effects of dif-

ferent forms of social support on prison staff job stress,

job involvement, job satisfaction, and organizational com-

mitment. Future researchers should consider using more

in-depth measures of job involvement, administrative sup-

port, and family support, which were measured in this

study with either 2 or 3 items. Future research should

examine theeffectsof other forms of socialsupport, such as

subordinate support as mentioned above. Another type of

social support not examined here was community support

(Gerstein, Topp, & Correll, 1987).

Research is needed to determine the best ways to

build social support among staff. For example, future

studies could examine whether teamwork and integra-

tion help build social support, or whether social support

plays a role in other outcome areas, such as work per-

formance or turnover/turnover intent. Additional studies

are needed to generate a more complete understanding of

the antecedents and effects of social support among prison

staff.

Two decades ago, Cullen (1994) urged that the concept

of social support be construed as an organizing concept

for the field of criminology. He saw social support as a

theme common to various theories of criminal behavior

and viewed it as a means of helping unify the field. Cullen’s

argument is that a lack of social support is criminogenic.

The present study builds on a limited body of past research

to demonstrate that social support in its various forms is

integral to understanding not only criminal action, but also

pertinent job outcomes among people working in prisons.

The findings presented here suggest that low levels of sup-

port are also associated with undesirable outcomes among

correctional staff. There is a kind of vicious cycle here,

in that staff who feel unsupported and experience high

job stress, low job involvement, low job satisfaction, and

low organizational commitment seem unlikely to provide

much support to the offenders in their charge. In this man-

ner, lowsupport could beget lowsupport, which, following

Cullen’s logic, might decrease the chances of successful

reentry of ex-offenders into the community.

The results of this study imply that correctional poli-

cymakers and leaders can promote better job outcomes

among their staff and improved effectiveness of their orga-

nizations by directly providingadministrative and supervi-

sory support to staff and also by making efforts to encour-

age peer coworkers to support one another. Such support

can be instrumental/goal oriented or expressive/affective

in nature, delivered by individuals or groups, and delivered

through either formal or informal channels (Cullen, 1994).

Likewise, the findings of this study imply giving the



10/11

http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0250http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0250http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0250http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0250http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0250http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0250http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0250http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0250http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0250http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0250http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0250http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0250http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0250http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0250http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0250http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0250http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0245http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0245http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0245http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0245http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0245http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0245http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0245http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0245http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0245http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0245http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0245http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0245http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0245http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0245http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0245http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0245http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0245http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0245http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0245http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0245http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0245http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0245http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0245http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0245http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0245http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0245http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0245http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0240http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0240http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0240http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0240http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0240http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0240http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0240http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0240http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0240http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0240http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0240http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0240http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0240http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0240http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0240http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0240http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0240http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0240http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0240http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0240http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0240http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0235http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0235http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0235http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0235http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0235http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0235http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0235http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0235http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0235http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0235http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0235http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0235http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0235http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0235http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0235http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0235http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0235http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0235http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0235http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0235http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0235http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0235http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0235http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0235http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0235http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0230http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0230http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0230http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0230http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0230http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0230http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0230http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0230http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0230http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0230http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0230http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0230http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0230http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0230http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0230http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0230http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0230http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0230http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0225http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0225http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0225http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0225http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0225http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0225http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0225http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0225http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0225http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0225http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0225http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0225http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0225http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0225http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0225http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0225http://link.springer.com/article/10.1007%2Fs10902-014-9522-7http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0215http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0215http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0215http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0215http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0215http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0215http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0215http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0215http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0215http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0215http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0215http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0215http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0215http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0215http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0215http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0215http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0215http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0215http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0215http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0210http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0210http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0210http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0210http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0210http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0210http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0210http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0210http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0210http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0210http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0210http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0210http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0210http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0210http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0210http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0210http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0210http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0210http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0210http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0210http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0210http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0205http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0205http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0205http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0205http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0205http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0205http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0205http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0205http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0205http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0205http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0200http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0200http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0200http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0200http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0200http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0200http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0200http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0200http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0200http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0200http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0200http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0200http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0200http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0200http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0200http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0200http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0195http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0195http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0195http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0195http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0195http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0195http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0195http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0195http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0195http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0195http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0195http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0195http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0195http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0195http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0195http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0195http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0195http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0195http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0195http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0195http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0195http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0195http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0195http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0195http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0190http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0190http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0190http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0190http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0190http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0190http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0190http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0190http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0190http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0190http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0190http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0190http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0190http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0190http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0190http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0190http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0190http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0190http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0190http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0190http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0190http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0190http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0190http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0190http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0190http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0190http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0185http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0185http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0185http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0185http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0185http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0185http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0185http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0185http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0185http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0185http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0185http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0185http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0185http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0185http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0185http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0185http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0180http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0180http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0180http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0180http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0180http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0180http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0180http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0180http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0180http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0180http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0180http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0180http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0180http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0180http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0180http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0180http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0180http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0180http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0180http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0180http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0175http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0175http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0175http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0175http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0175http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0175http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0175http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0175http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0175http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0175http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0175http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0175http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0175http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0175http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0175http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0175http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0175http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0175http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0175http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0175http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0170http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0170http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0170http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0170http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0170http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0170http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0170http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0170http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0170http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0170http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0165http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0165http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0165http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0165http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0165http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0165http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0165http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0165http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0165http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0165http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0165http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0165http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0165http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0165http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0165http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0165http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0165http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0165http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0165http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0165http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0165http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0165http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0165http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0165http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0160http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0160http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0160http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0160http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0160http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0160http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0160http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0160http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0160http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0160http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0160http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0160http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0160http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0160http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0160http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0160http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0160http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0160http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0160http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0160http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0160http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0160http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0155http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0155http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0155http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0155http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0155http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0155http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0155http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0155http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0155http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0155http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0155http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0155http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0155http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0155http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0155http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0155http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0155http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0155http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0155http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0155http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0155http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0155http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0155http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0155http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0155http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0155http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0155http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0155http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0155http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0150http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0150http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0150http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0150http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0150http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0150http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0150http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0150http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0150http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0150http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0150http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0150http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0150http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0150http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0150http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0150http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0150http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0150http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0150http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0150http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0150http://refhub.elsevier.com/S0362-3319(15)00085-3/sbref0150http://refhub.elsevier.com/S0362-

Documents

Social Support’s Relationship to Correctional Staff Job Stress,Job Involvement, Job Satisfaction, And Organizationalcommitment